With increasingly complex rendering and graphics processing operations requiring computationally heavy computing operations meeting memory bandwidth requirements can become challenging, as scaling memory bandwidth within a graphic processing device is non-trivial. In addition to, or as an alternative to increasing memory bandwidth within a graphics processing device, techniques have arisen to reduce the memory bandwidth requirements for certain operations and within certain components.